Therese Fuchs1, Andre Berger3, Dirk Fressmann4, Steffen Peldschus1,2

1 University of Munich, Germany, 2 Hochschule Furtwangen University, Germany, 3 ESI GmbH, Germany, 4 DYNAmore GmbH, Germany

Munich, 20th of April, 2016

Safeguarding a reliable validation and application of a FE human model in

different crash codes

2

Agenda

crash.tech | 2016/04/20

1. The THUMS User Community

2. Challenges in the validation of FE Human Models

3. Application of FE Human Models: pre- and post-processing examples

3

THUMS User Community (TUC)

Core Partners

Associated Partners

Subcontractor / Software Companies

crash.tech | 2016/04/20

4

Motivation of TUC

1. Harmonisation, provision and maintenance of a FE Human Body Model (THUMSTM)in the three crash codes LS-DYNA, VPS and Abaqus

2. Development of agreed procedures for the use of Human Body Models

o Guidelines for an improved model convertibility between codeso Development of validation procedureso Development of harmonised pre- and post-processing methods

crash.tech | 2016/04/20

5

Collaborations

crash.tech | 2016/04/20

COHERENT - project

6crash.tech | 2016/04/20

Validation of FE Human Models -Challenges

1. Choice of adequate reference experimentso Lack of physiological response at injurious levels of loadingo Ethical issueso Surrogate response variabilityo Choice of experimental boundary conditions (Suitable for numerical modelling?)o Documentation (“Old” reference?)o System-immanent uncertainties

2. Numerical realisationo Response variability due to choice of boundary conditions, output definition etc.o Harmonisation between codes (control settings, transformation, belt/airbag modelling)o Pre-simulation/Positioningo System-immanent uncertainties

3. Evaluation of validation resultso Filteringo Biofidelity ratingo Objective rating

7crash.tech | 2016/04/20

Body region Pelvis

Level Full Scale

Load case Lateral sled

ReferencesLeport et al. (2007):Assessment of the pubic force as a pelvic injury criterion in side impact. SAE Technical Paper, no. 2007-22-0019

Validation - Example

8crash.tech | 2016/04/20

28°

45°

45°

Initial position

Positioned

Validation - Example

9

Challenges in obtaining a harmonised position

Positioning of HBM required to match position of reference subjects

Two approaches to obtain harmonised positions in LS-Dyna and VPS:

o Apply same BCs in both codes for pre-simulation to position models independently

o Export nodal coordinates of positioned VPS model into LS-Dyna (or vice versa)

Problems with distorted elements in both cases

Manual mesh updates necessary in both codes after positioning

Validation –Harmonisation between codes

crash.tech | 2016/04/20

10

Application: Reference Points

crash.tech | 2016/04/20

45°

45°- PMHS positioned such that a 45° angle of the anterior pelvic

reference plane relative to horizontal plane

- Anterior pelvic plane is defined by the two anterosuperior iliacspines (ASIS) and pubic tubercles

11

Application: Rib Fracture Risk Assessment Tool

crash.tech | 2016/04/20

Method

- Strain-based probabilistic method to predict ribfracture risk using a material-level failure model

- Age-adjusted ultimate strain distribution used toestimate local rib fracture probabilities

- Strain-failure relationship based on probability thatlocal strains observed in ribcage would exceedultimate strain of rib cortical bone

o Estimation based on cumulative distribution ofrib cortical bone ultimate strains derived fromexperimental data from literature

o Kemper et al. (2005 & 2007): ultimate strainvalues obtained from uniaxial tensile testing ofrib cortical bone

Pipkorn, 2014 (5th International Symposium on Human Modeling and Simulation in Automotive Engineering)

12crash.tech | 2016/04/20

Tool

Application: Rib Fracture Risk Assessment Tool

13crash.tech | 2016/04/20

Application: Rib Fracture Risk Assessment Tool

14

Outlook

crash.tech | 2016/04/20

o Providing database of simulation setups for validation

o Development of harmonised pre-processing methods for the application of HBMs

o Post-processing: Development of objective criteria for the evaluation of the performance of HBMs, including improved tracking/reference points

15

THANK YOU!

Acknowledgment:

The work presented has been conducted by THUMS User Community, a project ofLMU in cooperation with Adam Opel AG, AUDI AG, Autoliv, BMW AG, Daimler AG,Dr. Ing. h.c.F. Porsche AG, Toyota Motor Corporation & Volkswagen Aktiengesellschaft.

Web link: www.TUC-project.org

crash.tech | 2016/04/20